You missed out another common reason: many reviewers feel that reviewing bad stuff (by their lights) isn’t worth their time or what they want to emphasize. They would prefer to alert readers to the better items “this is worth checking out.” And so very often reviewers have already done some leg work, having selected gear that has already stood out and impressed them in some set up.
There is no logic in that. How would they know something is bad without taking in to listen to in their place? How would they know how good a speaker is without listening to one that isn't? How would they learn anything if they don't experience bad designs?
There is only one reason: they want to keep getting (expensive) gear and no one would loan them such the first time they write a super negative review. They are not stupid.
If I was a reviewer and choosing my own subjects to review, I'd certainly choose those that were interesting in some way. The sonic results would be variable, of course. The extent to which "interesting" correlates with "good" would also vary, but that approach would weed out a fair bit of dross.
Any proper subjectivist reviewer would wait for the objective data in the case of Stereophile and if issues are found, analyze them subjectively. But they don't do that either even when their subjective impressions is at clear odds with the measurements.
I would listen, then measure and then listen again. And then EQ and listen
again. Which would be a combination of the Stereophile approach (listen before introducing measurement bias) and the ASR approach (measure, EQ, listen). I'd certainly want to listen fresh initially, then again to hear/confirm any audible correlation with measurements. And as I EQ to use a loudspeaker in practice, not doing so makes little sense for me. Hypothetically of course, as I'm not a reviewer. I wouldn't get many reviews done, I expect.
